Are modern computers the technological peak or will we see something in the future that will make them obsolete and outdated?
And no I'm not talking about the quantum meme because clearly that will never work (in b4 screeching, research it, it never worked and never will work). But is there something else on the horizon that I'm not aware of?
Other urls found in this thread:
Skynet is technological peak.
Can't be done because computers aren't aware of themselves and never can be. Linear output in linear output out, forever
Where there's a will, there's a way.
>Linear output in linear output out
Explain
You're assuming that computer design is entirely static, which is false. Quantum computers and various biological computer equivalent initiatives will eventually entirely change the design landscape of such things, in which case consciousness or intelligence could prove to be an emergent feature.
What do you think you brain is?
If you know anything about computers at all you should know they are nothing but machines that take some input, process it, and provide some output. How you make a procedural machine aware of itself is anyone's wild guess. If you have even 3% of brain power though, you can understand that it's completely impossible.
If you're still unconvinced, let this guy explain it to you in very simple terms:
youtube.com
tl;dr "true", cognitive AI is impossible
7 The appearance of the locusts was like horses prepared for battle; and on their heads appeared to be crowns like gold, and their faces were like the faces of men. 8 They had hair like the hair of women, and their teeth were like the teeth of lions. 9 They had breastplates like breastplates of iron; and the sound of their wings was like the sound of chariots, of many horses rushing to battle.
Revelation 9
Quantum computers don't work and never have. It is all one giant waste of money and something for researches to waste time on. It will never work. The output simply isn't stable. That's end of story. If I presented you with a CPU technology whose output is unstable, would you entertain me further? Would you fantasize about making it a reality? No, you would discard it.
A complex organic system created through millions of years evolution. If you consider the brain to be technology, then let me give you an analogy. If a UFO landed in your backyard and aliens popped out and just gave you the UFO so you could study it, you would have a much easier time understanding its principles than you could with the brain. The brain is "technology" on a level that we can't even begin to comprehend or formalize.
Your brain is a physical machine. Computers can in principle simulate all physical machines. Current lack of knowledge isn't evidence for the impossibility of knowledge.
>Computers can in principle simulate all physical machines
Do you have a couple trillions of years to spare? Because that's how long your "simulation" would take. "In principle."
Your brain fits in a skull sized volume. Even at a 1000x penalty to density such a machine could be practical. You're dodging the fundamental question because your claim of linear input and output is false.
Because the brain doesn't do linear operation. It doesn't take in one thing at a time and process it, it takes in everything at once and process everything.
>take some input, process it, and provide some output
So, like humans?
This thread is a fine example of how having just a little knowledge on a subject can make you think that you know a lot about the subject.
And in the event that you couldn't tell, i'm talking about you, OP.
Do you honestly believe that the x86-64 architecture is the peak of all possible computing? You're a fucking fool, it isn't even the peak of what we have right now.
Do a little more than read trendy tech websites and one day you'll understand.
If you knew even a little bit about neurochemistry and endocrinology you'd slap yourself back into diapers for what you just posted.
If you believe the laws of physics are computable then any turing machine can simulate any physical system. Therefore artificial intelligence is possible. The rest is just a question of optimization.
As far as minituarization goes, it's definitely closing in on a peak. Unless radically different technology gets invented this is what we're stuck on. Architecture is irrelevant, manufacturing and material limits is the real issue.
Holistic operation isn't something that complex user. The hardest thing to emulate in a brain is the ability to calculate wrong and accounting for how it "corrects itself" when recalling biased memories.
Does your brain not take care of everything at once? Please show me where exactly I'm in the wrong.
It's a question of the medium too. My claim is simply that our current technology is incapable of ever producing "true" AI. Now with some radically different technology it may be a different story.
Stochastic computing can get past physical limits on power efficiency. 3D stacking can multiply effective density hundreds or thousands of times. Cheaper chips caused by process stagnation means you can scale die area much further. That's all doable even if 7nm is the end of density.
Any parallel operation can be broken into a series of serial operations. Computers also don't take input serially and don't process serially. Your CPU has multiple cores and many I/O paths. Your GPU is a parallel processing monster.
Your CPU is bottlenecked by everything that isn't up to its speed and even then it's liable to have race conditions, and your GPU is a glorified matrix multiplier. Comparing it to the brain is not only laughable but also downright ridiculous. Try again.
Look up IBM Truenorth and Intel Loihi. Direct neural networks on chips. This is the same basic functionality the nervous system provides. It's already simulated well on GPUs and tensor processors. The new chips are for efficiency. Look up OpenWorm. Already existing simulation of simple animal intelligence. The tools are rudimentary but they exist.
A single-cell paramecium is still smarter than it. Good luck.
>Are modern computers the technological peak or will we see something in the future that will make them obsolete and outdated?
Did you read this and yet still proceed to post it?
Yes. Nothing will ever surpass what we have right now. We have reached our maximum technological potential and cannot go any further.
Asynchronous (clockless) chip design might be a very profitable avenue to go down. Intel did a research project that was an async version of the pentium mmx, and it was 3x as fast with half the power, iirc, on the same manufacturing process. It's apparently very difficult to verify logic designed in this manner, though.
I've illustrated the physical possibility and avenues for development. You have nothing to go on but doubt. I'm not saying these problems are easy but they are solvable. What do you think of the idea of growing biological accelerators for artificial intelligence? That's one of the possibilities I'm interested in.
It's fine as far as effort goes, it may even result in some discoveries and I'm all for that. But you should be realistic and serious about it and realize we'll never, ever make it behave intelligently.
Consider the fact that a paramecium is a single cell organism that is able to survive in the real world and breed. Now ask yourself just how impossibly advanced a computer chip would have to be to do the same? If you just threw it into the ocean, it would survive on its own and ensure future generations exist?
Yeah. But don't get me wrong, I'm all for advancing technology. It's just that people keep telling me how the AI is three days away and I automatically respond with every criticism I can think of when AI is mentioned. In other words, I'm an asshole.
>What do you think of the idea of growing biological accelerators for artificial intelligence?
It disgusts me frankly. We shouldn't be using tools whose principle of operation we can't even understand. Understanding first, implementation second. At least that's how I feel it should be done.
You can continue doubting but I believe it is unfounded. Current computation is so inefficient compared to what's possible even on the same manufacturing process it's comical. I don't see any reason to believe it's necessary to have an atomic level simulation of life to create intelligence. A large spiking neural network should be sufficient. Nothing is days away but progress will continue steadily through existing image processing, text prediction and game playing all the way to general intelligence. On the other side of the problem human cognitive enhancement through genetic engineering can expand what we are capable of manipulating feeding back into the full technologization of humanity. There's a man-machine fusion distant on the horizon.
Understanding how advancement is achieved is understanding the boundaries by which we are limited. So what are the boundaries? I ask because I am in no way an expert in this area.
Since your such a smartass, have you ever tried to reverse engineer something? Will it be a mechanical machinery, a chip or maybe something that evolution came up with? Apparently not because otherwise you wouldn't be talking so much shit, so here's a spoiler for ya: it's hard, really hard. The brain is ridiculously complex, so much that reverse engineer how it operates might be even impossible to do (without conducting ethically dubious experiments at least) and due to the way it physically changes over time it's already inferable that you cannot simply put it in a solid state chip. It's like to put together a schematic on how an engine is made based on the sounds it makes.
Please see video posted in For your own sanity's sake. AI only advanced in one direction, but the other direction is completely unexplored, and furthermore we don't even know how to approach it. There's just an enormous void of knowledge that we have no access to because no one's figured it out so far.
You can reverse engineer something you can vaguely understand. You can't - even vaguely - understand the brain. That's where your argument breaks down and becomes irrelevant.
>You can reverse engineer something you can vaguely understand
>...so here's a spoiler for ya: it's hard, really hard
Go away with your tardness if your reading comprehension skills are comparable to that of a orangutang
I see, you're going the ad hominem way because you feel insulted or something. Fine, be it your way.
Any argument for the technological impossibility of any physically possible task requires an argument for absolute technological stagnation of humanity. In my view the important question is whether there is enough remaining technological progress to reach either genetic engineering, general purpose nanotechnology, artificial intelligence, or simply artificial selection for more intelligent humans as a backup plan. All of those benchmarks expand the possibility for human technological expansion to limits we can't even begin to calculate. I think it's obvious that at least one of those possibilities will be fulfilled, most likely genetic engineering first.
HURR BRAIN IS THE MOST COMPLEX SHIT IN THE UNIVERSE BRO
t. brain with bloated ego
Nice defensive argument. I pointed you that reverse engineering is an hard job and MIGHT be impossible for the human brain given the current limitation on human experimentation yet you just went sperging "muh vaguely understanding is all you need"
Again, try to put together an engine by vaguely understaning how the sounds that is makes are puzzled together.
But that's correct?
According to the brain.
>Go away with your tardness if your reading comprehension skills are comparable to that of a orangutang
>defensive argument
Not responsing to a person with such fragile emotional balance, sorry.
The OP admitted that AI is physically possible. He only has doubts about the human capacity for implementing one. Considering the fact that technological development has continued steadily for the entire history of life on earth I would suggest that predicting an end to that trend is a poor judgement.
>not providing any counterargument
>misreading arguments provided
>falling back on muh emotions as last resort to have the last word
>pretend to be victorious
What's the next step?
Stop calling ad hominem an argument. Don't be surprised if I don't respond to you when your response to my perfectly civil post is
>Go away with your tardness if your reading comprehension skills are comparable to that of a orangutang
After that, I simply lose interest in anything you have to say because I'm not allowing you to drag me down to your level. Either present your arguments in a coherent, civil manner, or be ignored.
CRISPR up a thousand babies with every known positive gene for intelligence. Give them the highest quality care and education possible. Train them in all different disciplines. You now have a hyper Manhattan project for future technological development.
What if they refuse and instead take on gardening, painting, and other crafts?
If nature can create what we cannot in terms of complex and efficient systems, but we can manipulate nature, wouldn't that lead to a future where we're also possibly growing our chips instead of assembling them?
From a hardware standpoint, current hardware is more than sufficient for even your wildest compute needs. If one node isn't enough, you distribute and chain nodes. What the hell are you thinking about that needs more? You can get 32 cores on a desktop workstation ffs w/ tends of terabytes of storage. You seem to be speaking about hardware.
Tard #1
There are tons of exotic chips out there.
Plebs just seem to be unaware of them.
Brainlet gets informed by youtube e-celebs.
True AI is possible just not by the people brainlets follow. The average person is enamered by the bullshit they find on the web which is now-a-days 99% shekel lip service. They'll never understand anything beyond the brainlets they follow
> i'm a dumbass who knows nothing, therefore nothing complex or theoretical is possible
the post
>hurr I know everything
>but I'm too high and mighty to share my sources so just take my word on it okay?
Look everyone, an idiot
You might be a NEET but typically people do their jobs in society.
jews and other asians are the ones doing research now, so it's safe to say that we peaked already
OP hasn't posted any evidence that an artificial intelligence would even require the level of complexity in the human brain.